Stop Showing Me Mitch Keller's Percentile Chart


Remember to follow us on Twitter, Facebook, and Instagram for continued Pittsburgh Sports coverage: 

The Major League Baseball advanced analytics crowd is out of control.  They are becoming so enticed with these advanced analytics that they can’t believe what they are seeing on the field is true.  What was once used to explain trends is now being used as a barometer for player performance beyond the actual results played out on the diamond. 

I understand why advanced analytics became such a useful tool for those that are employed within the game.  It made Billy Beane one of the most well-known names in the history of MLB’s front office staff.  However, he was using it to supplement what he already knew about baseball that got him that job in the first place.

Now, Twitter is filled with these analytics “experts” that are trying to use advanced analytics to completely void results on the field.  They can’t realize that players do in fact succeed without being in the 90th percentile of Dip, XFip, XEra, and Bee-Bop-Do-Diddly-Doo, or they may be great in all of those and not succeed on the field.

Here’s what I mean.  When Trevor Williams had an outstanding, and unusually great, second half of 2018, Sabermetrics were correctly used.  It told us that since this was an outlier in Williams’ career, we should use advanced analytics to dive into it deeper.  Would it continue?  The numbers did indeed show that he was getting lucky, so to speak, and that trend would probably end.  Sure enough, it did. 

That is the correct way to use advanced analytics.  Using them to analyze an unusual trend to see if it will become the outlier or the new norm for a ballplayer.

The wrong way to use advanced analytics is to try to explain away consistent failure, or try to sh*t on successful players because they're falling behind in these advanced numbers.  The numbers are to supplement and analyze results, not overtake the results we can see with our own eyes.

Here are a few examples of the wrong way to use advanced analytics.  Mitch Keller is a bad pitcher.  He has shown us this with his career 6.08 ERA, and his continued inability to turn in two successful starts in a row.  Despite everyone getting in a tizzy over Keller showing an increased spin rate in a closed gym on turf somewhere, that went viral on Instagram, Mitch Keller has shown us, he is a bad pitcher.

However, there are still FanGraph fanboys that will cite all of these different numbers as to why Keller is, in fact, not a bad pitcher.  I won’t bore you with all of the acronyms or explanations to these stats.  If you’re a fan of sabermetrics, you know all of them anyway.  And yes, Keller is in good standing with some of the sabermetrics.  However, it doesn’t matter how “unlucky” you are or how much your spin rate has improved, if the results on the field are consistently poor. 

Now, if Keller were to all of a sudden turn it around, and start consistently turning in quality starts, then sure, explain to me how his improvement in velo and spin rate led to this new success.  Also, if those factors tell me it will continue, then great!  I love an explanation of a unique trend.  Or, if those numbers tell me this unique trend will not continue, as in the case of Williams, that is interesting to know as well. 

However, continued success or continued failure should not be countered with advanced analytics that lead to the contrary.  Mitch Keller is not a better pitcher than the results on the field because his XFip is above average.

And this isn’t a knock on FanGraphs.  I personally love FanGraphs.  I just don’t use it as a way to watch baseball instead of actually just, you know, watching baseball.

I’ve also seen Twitter “experts” concerned about Ke’Bryan Hayes and David Bednar.  I see that Hayes, “doesn’t barrel enough balls.”  Who cares?  The guy is batting .351 and shows no sign of slowing down.  My eyeballs tell me the guy is a ballplayer, and his consistent success on the field proves it.  He’s not an unusual trend. Hayes is a known good hitter.  We don’t need advanced analytics in this case.

The last straw was when I saw someone comment that they thought David Bednar’s Whiff percentage and Swing-And-Miss rate would be higher, and that “they are concerned about that.”  Bednar hasn’t had more than one earned run in an outing since restaurants and bars were at 50% capacity, and has had dozens of scoreless outings in the meantime.  I don’t care if his whiff percentage is zero.  The results speak for themselves. 

I am so tired of the FanGraphs fanboys trying to undermine other baseball fans by acting like they’re more knowledgeable solely because they can read a percentiles chart.  You can’t explain away Mitch Keller’s failure with a fancy graph.  You are not the smartest guy in the room just because you know what all of the acronyms stand for.  FanGraphs is a useful tool when used in the right way.  It is being used entirely the wrong way, and it has to stop.